Implement a new optimization in the inliner: if inlining multiple

calls into a function and if the calls bring in arrays, try to merge
them together to reduce stack size.  For example, in the testcase
we'd previously end up with 4 allocas, now we end up with 2 allocas.

As described in the comments, this is not really the ideal solution
to this problem, but it is surprisingly effective.  For example, on
176.gcc, we end up eliminating 67 arrays at "gccas" time and another
24 at "llvm-ld" time.

One piece of concern that I didn't look into: at -O0 -g with
forced inlining this will almost certainly result in worse debug
info.  I think this is acceptable though given that this is a case
of "debugging optimized code", and we don't want debug info to
prevent the optimizer from doing things anyway.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@80215 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 26 insertions, 0 deletions

diff --git a/test/Transforms/Inline/array_merge.ll b/test/Transforms/Inline/array_merge.ll
new file mode 100644
index 0000000000..f4f53ca1ce
--- /dev/null
+++ b/test/Transforms/Inline/array_merge.ll
@@ -0,0 +1,26 @@
+; RUN: llvm-as < %s | opt -inline | llvm-dis | FileCheck %s
+; rdar://7173846
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
+target triple = "i386-apple-darwin10.0"
+
+define internal void @foo() nounwind ssp {
+entry:
+  %A = alloca [100 x i32]
+  %B = alloca [100 x i32]
+  call void @bar([100 x i32]* %A, [100 x i32]* %B) nounwind
+  ret void
+}
+
+declare void @bar([100 x i32]*, [100 x i32]*)
+
+define void @test() nounwind ssp {
+entry:
+; CHECK: @test()
+; CHECK-NEXT: entry:
+; CHECK-NEXT: %A.i = alloca
+; CHECK-NEXT: %B.i = alloca
+; CHECK-NEXT: call void
+  call void @foo() nounwind
+  call void @foo() nounwind
+  ret void
+}